High-fidelity, high-resolution simulators, such as those used to train pilots, strive to immerse the participant in imagery of a virtual environment such that the simulator experience closely matches that of the real-world analog environment. Often, such a simulator may use an immersive display system that is made up of multiple display screens, or monitors, that form a dome, or other structure that at least partially encloses the participant. To ensure sufficiently high-resolution imagery, multiple image streams of the virtual environment may be generated, each of which may depict a particular portion of the virtual environment. Each image stream may then be displayed in a particular display area of the display system, such that the multiple image streams collectively form a relatively large, cohesive view of the virtual environment to the participant. Because of the processing requirements which may be necessary to generate high-resolution image streams, each individual image stream may be generated by a separate network-addressable image generation (IG) device that comprises a chassis, a processor, and a graphics subsystem capable of rendering the images that make up the corresponding image stream. Unfortunately, this may result in a large number of IG devices being necessary for the execution of a high-resolution and high-fidelity simulation. Such a large number of IG devices may require a substantial amount of physical space, and in many environments in which a high-fidelity simulator is used, space is at a premium. A large number of IG devices may also necessitate that substantial cooling apparatus be available in order to maintain the IG devices at a desired temperature. Accordingly, there is a need for mechanisms that allow the generation of multiple image streams of a virtual environment with fewer IG devices.